Climate intervention on a high-emissions pathway could delay but not prevent West Antarctic Ice Sheet demise

The paper addresses the ongoing debate surrounding solar radiation modification (SRM) as a potential tool to mitigate global warming and its associated risks, including ice sheet collapse. The effectiveness of SRM in halting anthropogenic climate change is supported by previous modelling studies. However, regional impacts and potential side effects remain highly uncertain. These uncertainties include impacts on the hydrological cycle, atmospheric chemistry, monsoon patterns, meridional temperature gradients, atmospheric and oceanic circulations, and modes of climate variability. Additionally, SRM does not address the direct effects of rising CO2 levels such as ocean acidification. The impact of SRM on polar ice sheets, particularly the WAIS, remains largely unexplored. Current research is limited to model simulations and laboratory studies, with field experiments facing societal and political obstacles. Ethical concerns such as delaying mitigation efforts, unilateral deployment, termination shock, and intergenerational injustice further complicate the issue. One argument for SRM is to prevent the crossing of climate tipping points, such as WAIS instability through self-sustained grounding-line retreat. The observed grounding-line retreat in West Antarctica is currently debated. A lack of progress in emissions reduction could lead to substantial ice-sheet loss and multi-meter sea-level rise on centennial timescales. Given these potential disruptions, the pressure to employ SRM might increase during the 21st century, even though there is very limited literature on its impact on polar ice sheets. The research aims to address this gap by studying the long-term response of the Antarctic Ice Sheet (AIS) under different emission and SRM scenarios.

The authors review existing literature on SRM, highlighting the uncertainties associated with its regional impacts and potential side effects. They note that while SRM's effectiveness in curbing global warming has been explored in several modeling studies, the specific effects on regional weather patterns and polar ice sheets remain largely unknown. The lack of robust understanding of potential side effects, coupled with the ethical and governance challenges surrounding SRM deployment, reinforces the need for the research presented in the paper. The review also covers existing studies on WAIS instability, noting the ongoing debate surrounding the causes and consequences of observed grounding line retreat. The authors emphasize the limited research on the impact of SRM on the AIS, specifically referencing the absence of studies employing explicit ice dynamic modeling. The paper positions its contribution within the gap in existing research, focusing on the long-term response of the AIS to geoengineering.

The study utilizes the Parallel Ice Sheet Model (PISM) with various parameterizations and grid resolutions (16 km, 8 km, and 4 km) to simulate the long-term response of the AIS under different scenarios. The model is forced with output from the HadGEM2-ES global coupled climate model, using three representative concentration pathways (RCPs): RCP 2.6 (low emissions), RCP 4.5 (intermediate emissions), and RCP 8.5 (high emissions). Four SRM scenarios (SRM85-20, -40, -60, -80) are considered for RCP 8.5, varying the starting year of stratospheric aerosol injections. The HadGEM2-ES simulations are extended to the year 2300, and the last 30 years of each simulation are repeated until the year 3000 to examine long-term impacts. The model includes parameters relevant for ice flow (till friction angle, sliding exponent) and basal ice-shelf melt (heat exchange coefficient), allowing for an ensemble of simulations to account for the uncertainties involved. Calving is represented using a heuristic relationship based on the horizontal strain-rate field. The model ensemble is calibrated against historical observational records of sea-level-equivalent ice volume, mass balance, annual sea-level contributions, ice thickness, and grounding line positions. The model spin-up includes three steps: a 200,000-year simulation under constant present-day forcing, a 2,000-year pre-industrial spin-up, and a run from 1860 to 2005 under common historical forcing. High-resolution simulations are restarted and regridded from the 16 km pre-industrial spin-up. For SRM scenarios starting later than 2020, the forcing is constructed by adding the climate anomalies from the SRM85-20 scenario to the corresponding RCP 8.5 climate state.

Under RCP 8.5 (high emissions), WAIS collapse is well underway by 2300, contributing 0.3–0.8 m to sea-level rise, with the entire AIS contributing 0.6–1.1 m. By 3000, WAIS contributes 2.1–2.8 m and the AIS contributes 2.9–4.0 m to sea-level rise. The timing of Thwaites Glacier collapse varies with model resolution. At 16 km resolution, the collapse of the Ross Ice Shelf precedes Thwaites retreat, suggesting a drainage-basin-scale interaction. At higher resolutions, Thwaites retreat starts earlier. The East Antarctic Ice Sheet remains largely stable until 3000. The impact of climate change on Antarctic ice-mass change becomes discernible only in the second half of the 21st century. RCP 2.6 (rapid decarbonization) and early SRM interventions (starting in 2020 or 2040) are the only scenarios that mostly ensure WAIS stability. However, even in RCP 2.6 and early SRM scenarios, committed ocean warming could still trigger WAIS collapse in some model runs where Thwaites Glacier is already near a tipping point. Delaying SRM interventions under RCP 8.5 increases the probability of WAIS collapse and the magnitude of sea-level rise. In RCP 4.5 (intermediate emissions), mid-century SRM interventions have a substantial risk of failure in preventing WAIS collapse. The study shows that the longer action is delayed (emissions mitigation or SRM), the more likely the WAIS collapse becomes. The analysis reveals that even with strong mitigation (RCP 2.6) or early SRM (SRM85-20), a committed warming of the Southern Ocean continues, which can still be sufficient to trigger WAIS collapse in some simulations, given the possibility of Thwaites Glacier already being near a tipping point. This suggests that the effectiveness of SRM intervention is highly sensitive to timing and the initial state of the ice sheet.

The findings highlight the importance of both rapid decarbonization and early deployment of planetary-scale SRM to prevent WAIS collapse. However, the paper stresses the political challenges, potential side effects, and uncertainties associated with SRM. The study's conservative model setup suggests the presented results represent a lower bound of the challenge faced. More sensitive models might show an even lower effectiveness of SRM interventions. The reliance on a stand-alone ice-sheet model, neglecting feedbacks from ice-sheet topography changes and freshwater input, might underestimate ice-sheet loss. The model resolution also affects the results, with higher resolutions indicating less effective SRM interventions. The research underscores the need for a better understanding of the regional impacts of SRM before large-scale deployment. The authors reinforce the crucial role of significant GHG emission reductions in ensuring long-term WAIS stability. While SRM might offer a limited time window for adaptation, it cannot replace emissions reduction as the primary strategy.

The study concludes that rapid decarbonization is the most effective way to prevent long-term WAIS collapse. While early deployment of planetary SRM could offer a supplementary measure, particularly in intermediate-emissions scenarios, its associated risks and uncertainties necessitate a cautious approach. The significant uncertainties in regional impacts of SRM interventions emphasize that emissions reductions remain the crucial strategy for preserving WAIS stability. Future research should focus on coupled ice-sheet-climate models to better account for complex feedback mechanisms and improve estimates of SRM's effectiveness.

The study employs a stand-alone ice-sheet model, neglecting feedback mechanisms between the ice sheet and the atmosphere and ocean. This simplification might underestimate the rate of ice sheet loss. The model's resolution and initial configuration might also influence the results, suggesting a conservative estimate of SRM's efficacy. The reliance on a specific climate model (HadGEM2-ES) and its representation of Antarctic climate may limit the generalizability of the findings. Moreover, the analysis does not consider the possibility of actively reversing temperature conditions through SRM after a climate threshold is exceeded, which could offer a potential, though uncertain, mitigation strategy.